XXVII. Cold Gas and the Colours and Ages of Early-Type Galaxies
Total Page:16
File Type:pdf, Size:1020Kb
MNRAS 444, 3408–3426 (2014) doi:10.1093/mnras/stt2474 The ATLAS3D project – XXVII. Cold gas and the colours and ages of early-type galaxies Lisa M. Young,1,2‹ Nicholas Scott,3 Paolo Serra,4,5 Katherine Alatalo,6 Estelle Bayet,7 Leo Blitz,8 Maxime Bois,9 Fred´ eric´ Bournaud,10 Martin Bureau,7 Alison F. Crocker,11 Michele Cappellari,7 Roger L. Davies,7 Timothy A. Davis,12 P. T. de Zeeuw,12,13 Pierre-Alain Duc,10 Eric Emsellem,12,14 Sadegh Khochfar,15 Davor Krajnovic,´ 16 Harald Kuntschner,12 Richard M. McDermid,17,18,19 4,20 21 4,20 22 Raffaella Morganti, Thorsten Naab, Tom Oosterloo, Marc Sarzi Downloaded from and Anne-Marie Weijmans23 Affiliations are listed at the end of the paper Accepted 2013 December 16. Received 2013 December 11; in original form 2013 November 5 http://mnras.oxfordjournals.org/ ABSTRACT 3D We present a study of the cold gas contents of the ATLAS early-type galaxies, in the context of their optical colours, near-ultraviolet colours and Hβ absorption line strengths. Early-type (elliptical and lenticular) galaxies are not as gas poor as previously thought, and at least 40 per cent of local early-type galaxies are now known to contain molecular and/or atomic gas. This cold gas offers the opportunity to study recent galaxy evolution through the processes at European Southern Observatory on October 27, 2015 of cold gas acquisition, consumption (star formation) and removal. Molecular and atomic gas detection rates range from 10 to 34 per cent in red sequence early-type galaxies, depending on how the red sequence is defined, and from 50 to 70 per cent in blue early-type galaxies. Notably, massive red sequence early-type galaxies (stellar masses >5 × 1010 M, derived from dynamical models) are found to have H I masses up to M(H I)/M∗ ∼ 0.06 and H2 masses up to M(H2)/M∗ ∼ 0.01. Some 20 per cent of all massive early-type galaxies may have retained atomic and/or molecular gas through their transition to the red sequence. However, kinematic and metallicity signatures of external gas accretion (either from satellite galaxies or the intergalactic medium) are also common, particularly at stellar masses ≤5 × 1010 M, where such signatures are found in ∼50 per cent of H2-rich early-type galaxies. Our data are thus consistent with a scenario in which fast rotator early-type galaxies are quenched former spiral galaxies which have undergone some bulge growth processes, and in addition, some of them also experience cold gas accretion which can initiate a period of modest star formation activity. We discuss implications for the interpretation of colour–magnitude diagrams. Key words: galaxies: elliptical and lenticular, cD – galaxies: evolution – galaxies: ISM – galaxies: structure – radio lines: galaxies. mid-IR polycyclic aromatic hydrocarbon (PAH) emission, optical 1 INTRODUCTION emission lines and cm-wave radio continuum emission (e.g. Sarzi While early-type (elliptical and lenticular) galaxies generally have et al. 2010; Shapiro et al. 2010). Surveys of UV and optical colours smaller relative amounts of star formation activity than spirals, they have suggested that this kind of low-level star formation activity are not all completely devoid of such activity. Modest rates of star is present in around a quarter of all nearby early-type galaxies formation in early-type galaxies can be traced with ultraviolet (UV) (Yi et al. 2005; Kaviraj et al. 2007; Suh et al. 2010). Recently colours, mid-infrared (IR) and far-IR (FIR) continuum emission, Smith et al. (2012) have emphasized that star formation can be found even in early-type galaxies which are on the red sequence, and Fumagalli et al. (2013) have measured star formation rates in E-mail: [email protected] optically quiescent galaxies out to redshifts of 1–2. C 2014 The Authors Published by Oxford University Press on behalf of the Royal Astronomical Society Cold gas in early-type galaxies 3409 This evidence for star formation activity should be understood in type sample is actually drawn from a parent sample which has no the context of paradigms for the development of the red sequence. colour or morphological selection, and optical images of the entire Star formation requires cold gas, and the treatment of cold gas is parent sample have been inspected by eye for large-scale spiral a crucial element of galaxy evolution models. For example, one structure. The 260 galaxies lacking spiral structure form the basis 3D class of models assumes that the development of the red sequence of the ATLAS project; integral-field optical spectroscopy over a field is driven by active galactic nucleus (AGN) activity, as AGN-driven of at least 33 arcsec × 41 arcsec was obtained with the SAURON outflows remove the star-forming gas from blue gas-rich galaxies instrument (Bacon et al. 2001) on the William Herschel Telescope (e.g. Kaviraj et al. 2011). The outflow is assumed to remove the for these 260 galaxies. Paper I also tabulates their assumed distances gas more quickly than it could be consumed by star formation and foreground optical extinctions. 3D activity.Coldgasmayalsoberemovedorconsumedbyavarietyof The ATLAS early-type galaxies clearly trace out the optical red other processes associated with bulge growth (Cheung et al. 2012). sequence, and they also include a smaller population of somewhat Another class of models points out that it may not be necessary to bluer galaxies (Paper I). Detailed study of the stellar kinematic maps remove the cold gas, but simply to sterilize it (Kawata, Cen & Ho is used to reveal internal substructures such as counter-rotating 2007; Martig et al. 2009, 2013). In order to test these paradigms stellar cores or kinematically decoupled stellar discs (Krajnovic´ for the development of the red sequence, and in particular to test et al. 2011, hereafter Paper II). The specific angular momenta of Downloaded from the relative importance of gas recycling, accretion and removal the galaxies are analysed in Emsellem et al. (2011, hereafter Paper as galaxies move to and from the red sequence, we explore the III) in the context of formation paradigms for slow and fast rotators relationships between the colours of early-type galaxies, their Hβ (Bois et al. 2011; Khochfar et al. 2011; Naab et al. 2014). Dynamical absorption line strengths and their cold gas content. masses are provided by Cappellari et al. (2013a, 2013b). Other work There have been several other surveys for molecular gas in var- on the sample probes environmental drivers of galaxy evolution ious samples of early-type galaxies (e.g. Welch, Sage & Young (Cappellari et al. 2011b; Cappellari 2013), hot gas (Sarzi et al. http://mnras.oxfordjournals.org/ 3D 2010), but the ATLAS project is the only one which also has the 2013), stellar populations and star formation histories (McDermid kinematic information that is necessary to interpret galaxies’ recent et al., in preparation), and bulge/disc decompositions (Krajnovic´ evolutionary histories. Stellar kinematics, shells and tidal features et al. 2013). give insight into the merger and assembly histories of the galax- ies, and ionized gas, H I, and molecular kinematics reveal signs of 3D gas accretion and/or gravitational disturbances. Thus, the ATLAS 3DATA project is the first opportunity to bring together the evidence from star formation activity (recorded in the colours and Hβ absorption of 3.1 Cold gas, line strengths and stellar population ages the stellar populations) and the recent interaction/accretion history 3D at European Southern Observatory on October 27, 2015 All of the ATLAS early-type galaxies except NGC 4486A were ob- recorded in the gas. In this paper we do not deal with measurements served with the institut de radioastronomie millimetrique (IRAM) of an instantaneous star formation rate or star formation efficien- 30 m telescope in the 12CO J = 1−0and2−1 lines (Young et al. cies derived therefrom (e.g. Martig et al. 2013). Instead, we focus 2011, hereafter Paper IV).1 The CO detection limits correspond to particularly on the connections between gas accretion and the star H masses of approximately 107 M for the nearest galaxies and formation history over the last few Gyr, as those are encoded in the 2 108 M for the more distant galaxies. To place this CO survey in stellar populations. 3D context it is crucial to note that the luminosity selection criterion for Section 2 of this paper describes the ATLAS sample of local the sample is based purely on the total K -band stellar luminosity, early-type galaxies. Section 3 gives explanatory information on S not the FIR flux or even the B luminosity (which have been used the provenance of the colour, stellar population and cold gas data. in the past, and which are strongly influenced by the presence of Section 4 shows the atomic and molecular gas contents of early- star formation). The CO detection rate is 22 ± 3 per cent; detected type galaxies, both on and off the red sequence. Depending on which 7.1 9.3 ± ± H2 masses range from 10 to 10 M and the molecular/stellar colours are used to define the red sequence, 10 2to34 6per −4 mass ratios M(H )/M∗ range from 3.4 × 10 to 0.076. In addition, cent of red sequence early-type galaxies have >107 M of cold 2 the CO detection rate and M(H )/M∗ distributions are broadly con- gas. A discussion of internal extinction is found in Section 5, which 2 sistent between Virgo Cluster and field galaxies, so that the cluster shows that although the H -rich galaxies are dusty, the dust usually 2 environment has not strongly affected the molecular gas content.